The primary purpose of the project on fertility is to develop methods for identifying reproductive hazards. Ultimately, we are working toward developing methods for studying reproductive hazards in small populations so that localized exposures can be evaluated. First, we have developed questionnaire methods for measuring fertility using data on time to pregnancy, the number of noncontracepting menstrual cycles required to conceive. This measure can be the focus of epidemiologic studies of fertility or can be integrated into epidemiologic studies of other diseases such as breast cancer. We applied this method in a study of mercury vapor and nitrous oxide exposures in dental assistants. In addition to studying the effect of these exposures on fertility, we evaluated the accuracy of time-to- pregnancy data from a short, self-administered questionnaire. We also have used this approach to study the effects of prenatal exposure to DES. Immunological impairments associated with prenatal DES exposure were also examined (manuscript, in press). We are using the DES study to examine the quality of time-to-pregnancy data provided by male respondents. Our second approach focusses on developing biological markers of impaired ovarian function. We have begun by examining urinary metabolites of estrogen and progesterone. These have been measured in daily urine specimens from over 700 menstrual cycles from about 200 women. First, we developed a method for detecting day of ovulation using these urinary hormone data. Our ability to detect ovulation using this method was very good (described in manuscript in press in Epidemiology). Now we are developing methods for characterizing follicular development and luteal function.